Moving Higher Education AI from Pilots to Strategy

Higher education is at an inflection point. Ninety-four percent of higher education professionals are already using artificial intelligence (AI) in some capacity.^Footnote1 Yet many institutional policies, governance structures, and infrastructure haven't kept pace. This gap is where institutions use AI to drive transformation or find themselves managed by it.

This tension was at the center of a panel I moderated at the 2026 EDUCAUSE Annual Conference. Panelists included Shelley Knuth, assistant vice chancellor for Research Computing at the University of Colorado Boulder, Stan Waddell, vice president and CIO at Carnegie Mellon University, and Louis Stewart, head of ecosystem development at NVIDIA. The conversation was candid, practical, and urgent—focused on what's actually happening on campuses, not just what's being planned. As Waddell explained, "We have moved away from 'Should we be doing it?' to 'How do we incorporate this very disruptive technology into pedagogy, into operations, and how do we move away from the paradigms that really got us started?'"^Footnote2

Shifting from Grassroots Pilots to Institutional Scale

Across every campus represented in our discussion, the same dynamic is playing out: AI experimentation has grown faster than institutions can govern it. Waddell framed this not as "shadow IT" but as a "citizen-led development era." The goal isn't to stop AI experimentation. It's to build the structures that can evaluate it, scale what works, and sunset what doesn't.

Carnegie Mellon is doing exactly that: identifying the top 10–20 percent of pilots that demonstrate real value and professionalizing them into repeatable, enterprise-grade services.^Footnote3 The institution has also merged data governance and AI governance into a single advisory committee, recognizing that the people responsible for one are often responsible for the other. At Dell, we've found the same to be true: effective governance that safeguards data, privacy, and trust provides guardrails for AI without slowing down innovation.

Knuth pushed the thinking further, challenging institutions to reorient not just their governance and data structures but their investment decisions. "The question has been around 'What do we need to do to build AI?'" she said. "And it needs to start shifting toward 'What's the problem we're trying to solve, and what do we need to solve that problem where AI can help us?'"^Footnote4

Managing Infrastructure Costs and Supporting Student Success

Getting AI strategy right requires good governance, data readiness, scalable infrastructure, and creative solutions given tight budgets. Knuth described how rising hardware costs are pushing institutions toward shared models, such as the Rocky Mountain Advanced Computing Consortium, where institutions pool resources rather than build independently.^Footnote5 That's a model worth examining at every institution navigating budget constraints.

The conversation surfaced another fundamental challenge: Higher education has never had enough advisors to give every student the personalized support they need. What AI offers is something institutions have never had before: the ability to deliver personalized, responsive support at scale—twenty-four hours a day and for every student, not only for those who know how to ask for help. Waddell noted that institutions deploying virtual advisors around the clock are already seeing beneficial outcomes. Students in crisis get help when they need it, retention improves, and fewer students fall through the cracks.^Footnote6 That's not incremental progress. It's a fundamentally different kind of institution.

Measuring What Actually Matters

Stewart discussed his "triple helix" model—aligning government, industry, and academia around shared infrastructure goals—as a framework for extending AI benefits beyond flagship universities to community colleges, rural communities, and underserved populations.^Footnote7 But realizing that potential requires institutions to measure what really matters.

Return on investment (ROI) has always been a complicated metric in higher education, and AI doesn't simplify it. Waddell acknowledged that when AI automates a significant portion of a role in the private sector, that position is often eliminated.^Footnote8 In higher education, that capacity tends to get redirected toward other priorities, which is the right call for mission-driven institutions but makes traditional ROI harder to demonstrate.

This is why the idea of "return on mission" is so compelling. The metrics that matter in higher education are student retention and graduation rates, reduced crisis incidents, faster research timelines for greater breakthroughs, and higher satisfaction among faculty and students. These outcomes are what mission looks like in practice.

Before your next board or budget conversation, consider whether you can connect your AI investments directly to the outcomes your institution was built to deliver.

Understanding What This Means for Institutions

The institutions making the most progress on AI strategy are disciplined and focused on where AI solves real problems rather than chasing every new capability. They're investing in data foundations before scaling new tools, building cross-functional governance processes, and measuring success in mission terms rather than balance-sheet terms. And they're not trying to out-recruit the private sector for AI talent. They're building it from within, upskilling the faculty and staff they already have.

Perhaps the clearest lesson from the panel discussion is that AI strategy cannot be owned by a single office. Researchers, administrators, faculty, students, and technology leaders all have a role to play in determining where AI creates value and how it should be governed.

Taking the Next Step

The path from AI experimentation to AI at scale requires clarity of purpose, shared governance, and infrastructure that can support what's coming.

At Dell, we're working alongside higher education institutions at every stage of the AI journey. Whether evaluating potential use cases, building consensus across campus stakeholders, identifying funding opportunities, or developing a roadmap for responsible AI adoption, we are here to help.

Thank You

Notes

1. Jenay Robert, The Impact of AI on Work in Higher Education (EDUCAUSE, January 2026). Jump back to footnote 1 in the text.↩
2. JJ Davis et al., "From AI Experimentation to Institutional Strategy: What Higher Education Leaders Are Learning Now," EDUCAUSE Annual Conference, Denver, CO, September 30, 2026. Jump back to footnote 2 in the text.↩
3. Ibid. Jump back to footnote 3 in the text.↩
4. Ibid. Jump back to footnote 4 in the text.↩
5. Ibid. Jump back to footnote 5 in the text.↩
6. Ibid. Jump back to footnote 6 in the text.↩
7. Ibid. Jump back to footnote 7 in the text.↩
8. Ibid. Jump back to footnote 8 in the text.↩

Jennifer "JJ" Davis is Senior Vice President, Americas Marketing, at Dell Technologies.